ORIGINAL ARTICLES
Bromide Treatment of Pharmaco-Resistant Epilepsies with Generalized Tonic-Clonic Seizures: A Clinical Study Bernhard J Steinhoff, MD and Rolf Kruse, MD
In a retrospective controlled clinical study we investigated the efficacy of bromides (BR) as to the frequency of generalized tonic-clonic seizures (GTCS) as a mono therapy (n = 5) or combined with other antiepileptic drugs (AEDs) (n = 55), and compared the results statistically with those achieved with a treatment based on phenobarbitone (PB) and/or phenytoin (PHT) in a very similar group of another 60 patients. All patients treated with BR had previously proved to be pharmaco-resistant. In most of the cases the epilepsies were based on early cerebral lesions, and were partly associated with severe mental and physical handicaps. Most of the patients were between three and 14 years old. The percentage of responders was 58%, which means a reduction in the GTCS-frequency of more than 50%. The positive result was observed for 28 months on the average. 27% of all patients became free ofGTCS for 18 months on the average (range, 44 days to 62 months), and 32% were improved (reduction in GTCS-frequency of more than 50% while not GTCS-free) for nearly three years on the average (range, 79 days to 110 months). Compared to PHT or PB, BR were slightly more efficient and similarly tolerable, but no statistically significant difference was seen (48% responders in the control group versus 58% responders in the BR-group). Thus, the statistical co"elation with competitive therapies confirmed the good efficacy of BR against GTCS, which was previously demonstrated by clinical tests without control studies. We recommended BR in the case of every pharmaco-resistant epilepsy with GTCS as a therapeutic alternative. Key words: Epilepsy, pharmaco-resistance, treatment, bromides, phenobarbitone, phenytoin, serum concentrations. Steinhoff BJ, Kruse R. Bromide treatment of pharmaco-resistant epilepsies with generalized tonic-clonic seizures: a clinical study. BrainDev 1992;14:144-9
The bromides (BR), introduced as antiepileptic drugs (AEDs) by Locock in 1857 [1], had lost their clinical importance much more since the appearance of phenobarbitone (PB) in 1912, phenytoin (PHT) in 1938 and then modern first-line AEDs [2]. However, their efficacy, especially in the case of generalized tonic-clonic seizures (GTCS), has been emphasized in several, though noncontrolled and statistically not satisfactory, investigations and reports [3-11]. Still in 1940, for instance, Southerland [12] reported that 85% of 40 patients treated with BR in addition to PB had less seizures than with PB alone,
From the Department of Children and Adolescents, Epilepsy Center Kork, Kehl-Kork. Received for publication: December 3, 1991. Accepted for publication: February 6,1992. Correspondence address: Dr. med. Bernhard J. Steinhoff, Neurologische Klinik, Ludwig-Maximi1ians-Universitiit, Marchioninistr. 15, D-8000 Miinchen 70, Germany.
and thus recommended BR as a very effective add-onAED. For 55 years BR had been the only effective AEDs. Therefore their antiepileptic indication was unlimited. Later, BR were recommended for the treatment of therapy-resistant epilepsies with GTCS in children and adolescents by only a few authors [13-18]. Livingston [17] and Berman [13] recommended them also for myoclonic seizures, and Matthes [19] for the treatment of partial seizures. Others [20] used BR only in cases of epilepsy with porphyria. Nevertheless, in some specialized epilepsy centers dealing with intractable epilepsies, BR were still used as effective add-on-drugs in selected cases, mainly with GTCS [2]. Recently, the clinical usefulness of BR was re-confirmed by several authors [21-26]. However, a statistical comparison of the clinical efficacy of BR with results achieved with established first-line AED has never been reported.
We report the results of a retrospective and controlled study concerning the efficacy of long-tenu BR therapy in 60 patients with pharmaco-resistant GTCS to conventional AED treatment.
Table 1 Clinical data
The efficacy of long-term BR therapy was investigated in 60 patients with pharmaco-resistant GTCS to conventional AEDs between 1972 and 1985. The control group also comprised 60 patients with GTCS treated with PB and/or PHT as a basic medication. For both the BR group and the controls the following selection criteria were used: 1. Epilepsies with either GTCS alone or combined with other seizure types. 2. Mean frequency of GTCS of two or more seizures per month for the two years before starting the treatment or from the very beginning of epileptic seizures in cases of a shorter duration than two years. 3. No or only slight changes in current antiepileptic co-medication. 4. No additionally introduced AEDs during the observation period. 5. Treatment only in Epilepsy Centre Kork. The definition of the patients with pharmaco-resistant GTCS to conventional AED treatment is those sufferred from severe epilepsies with GTCS which were insufficiently controlled by treatment with conventional AEDs like PB, primidone (PRM), PHT, carbamazepine (CBZ) or valproic acid (VPA) at the highest tolerable doses by Epilepsy Centre Kork. The clinical data are shown in Table 1. Both groups contained equal numbers of male and female patients, however, there were some, but not considerable, differences with respect to the mean ages of onset of epilepsy and of treatment, in addition to the mean duration of epilepsy, and to the relationship of pure and combined GTCS epilepsies, but not with respect to epileptic syndromes defined according to the proposal on the classification of epilepsies and epileptic syndromes of the International League against Epilepsy (ILAE) [27], and not with respect to etiological factors or to neurological and mental defects. In both groups the epilepsies were mostly severe residual syndromes based on early acquired cerebral lesions. In nearly all cases the epilepsy started in the first two years of life. Mental and neurological handicaps as well as pathological neuroradiological findings were commonly seen in these two groups. The group of generalized epilepsies consisted of symptomatic forms or both symptomatic and/or idiopathic fonus (groups 2.2 and 2.3 in the ILAE classification) including Lennox-Gastaut syndrome. Two patients in the control group and only one in the BR group had an
Control group
30/30 1.3 (0.3-11.3) (2.2) 10.0 (1.5-27.6) (5.6) 8.9 (0.6-25.9) (5.5)
30/30 2.3 (0.2-12.4) (3.0) 8.9 (0.5-18.1) (4.1) 6.3 (0.1-17.7) (4.2)
20 40 21
18 42 9
49
46 2 12 21 39
n = 60
Gender male/female Mean age at onset of epilepsy Range (years)
PATIENTS AND METHODS
Bromide group
(s)
Mean age at onset of treatment Range (years) (s)
Mean duration of epilepsy Range (years) (s)
Epilepsy syndrome Generalized Undetermined Epilepsy with GTCS alone Etiology Residual Progressive * Unknown Neurological deficits Mental defects Monotherapy
26 44 5
Combined therapy
55
11
n = 60
PB PHT PB PHT PB+PHT
4 5 18 17 16
* In both cases tuberous sclerosis, GTCS: generalized tonic-clonic seizures, PB: phenobarbitone, PHT: phenytoin. idiopathic generalized epilepsy (group 2.1. in the classification), and there was no case of typical localizationrelated symptomatic or idiopathic epilepsy (group 1 in the classification). The majority had to be classified as undetermined epilepsies (group 3 in the claSSification) with both generalized and focal signs, and seizures like generalized epilepsies beginning as idiopathic or symptomatic epilepsies with GTCS, but later in the course with partial seizures like complex-partial seizures additionally and with corresponding focal abnormalities in the EEG. There were no cases of severe myoclonic epilepsy in infancy, ESES syndrome or Landau-Kleffner syndrome in either group. If GTCS was not the only seizure type, there were additional tonic, minor generalized and localization-related seizures in different combinations and frequencies. Five patients were treated with BR monotherapy, while five patients were only on PHT and four on PB in the control group. BR were combined with a single AED (PB, PRM, PHT, CBZ, or VP A) in 20 cases and with two or more other AEDs in 35 cases, reflecting the severity of these epilepsies with GTCS of high frequency. 57 patients received KBr and the others CaBr2. Only BR serum concentration (SC) values determined by x-ray fluorescence analysis were examined. This
Steinhoff and Kruse: Bromide therapy in intractable
ercs
145
Table 2 Success rate of bromide treatment Effect Free of GTCS Improvement (n) >50% (n)
Duration of therapy effect > > > >
1.5 months 3 months 1 year 3 years
16 12 6 3
Responders (%)
Total no of patients
58% 55% 59% 67%
60 55 32 15
19 18
13 7
GTCS: generalized tonic-clonic seizures, improvement> 50%: reduction in GTCS-frequency > 50%, responders: free of GTCS/ improvement> 50%.
Table 3 Age at onset of bromide therpay and success rate (n = 60) Age (years)
n
FreeofGTCS (n)
Improvement >50% (n)
Responders
14
13
3
(%)
54%] 31%
46%*
* p < 0.05, GTCS: generalized tonic-clonic seizures, improvement> 50%: reduction in GTCS-frequency > 50%, responders: free of GTCS/improvement > 50%.
method, described elsewhere in detail [17, 34], was applied in 49 cases. The aim of the present study was to determine the therapeutical effects on the GTCS-frequency and the side effects of the BR-treatment compared with the control therapies. The effectiveness was classified as follows: 1. Free of GTCS for at least 1 month, i.e., the two fold time of being seizure-free before starting the treatment with BR or with PHT or PB in the control group. 2. Improvement, i.e., a reduction in GTCS-frequency by more than 50%, for at least the same time. These two subgroups are also summarized as "responders." 3. No effect. 4. Increase in GTCS-frequency, i.e., worsening. We compared the results achieved in both groups by means of the chi 2 -test, modified according to Yates (p = 0.05) [2]. RESULTS Table 2 shows that 16 patients treated with BR became free of GTCS and 19 patients improved remarkably. Therefore, 58% of the patients were responders for at least six weeks, when they had attained the BR steady state. This duration of at least six weeks is three fold
146 Brain & Development, Vol 14, No 3, 1992
longer than the response duration which had been attained during the two years before the BR treatment was started. Beside this initial positive effect the mean response duration was remarkably long: Patients, who had become GTCS-free, remained seizure-free for 18 months on average (44 days to 62 months). Improvement lasted even for 36 months on average (79 days to 110 months). In the follow-up period the success rate remained between 55% and 67% (Table 2). Thus, a loss of efficacy was not experienced. In one third (n = 12) of the 35 responders GTCS reappeared at the frequency seen in the pretreatment period. However, if the positive effect was maintained for longer than one year, we noticed a relapse of the pretreatment GTCS-frequency in only four of the remaining 19 responders. Many correlations were examined, i.e., between the success rate and different parameters of supposed prognostic importance, but only a few correlations were of statistical significance, for instance, the correlation of the success rate and age at the onset of BR therapy (Table 3). The treatment of patients younger than seven years, who showed more generalized epilepsies than average, proved significantly more often to be successful than the treatment of older patients. Although we did not find any statistically significant correlation between the duration of epilepsy and the success rate, it is remarkable that a positive therapeutical effect, sometimes for years, was obtained in half the 41 patients with a duration of epilepsy of more than five years (n = 22) and in 1/3 of the 25 cases with epilepsy for more than ten years (n = 9). With respect to the classification of the epileptic syndromes [27], a positive therapeutical effect could be obtained only in 55% of the cases of undetermined epilepsies (n = 22), compared to 65% of those of generalized epilepsies (n = 13). However, the rate of a positive long-term effect for more than one year was significantly higher in the cases of undetermined epilepsies (n = 16 (73%) versus n = 3 (23%), p 1.5 months
48% NS
Bromide group Control group
60 60
b. Success rate after> 1 year
66%
Bromide group Control group
Responders: free of generalized tonic-clonic seizures or improvement> 50%, NS: not significant (p < 0.05).
32 29
Responders: free of generalized tonic-clonic seizures or improvement> 50%, NS: not significant (p = 0.05).
30 Number of patients (n ~ , 201
2 ·· .. ·3·
free of GTCS -
improvement **
Bromide Therapy
(n = 60)
I!l!II
no effect
wo rsening
Control Therapy
(n = 60)
Fig 1 Efficacy as the CTCS* in bromide and control groups. *generalized tonic·c1onic seizures, **reduction in CTCSfrequency> 50%.
The mean SC determined in 49 patients was 253 mg/dl (156 mg/dl to 367 mg/dl). More than 70% of the measured mean SCs were between 200 and 300 mg/dl. The incidence of responders with mean SCs higher than 250 mg/dl was higher than with lower SCs, but again the difference was not significant (Table 4). We found a weak correlation between SC and the mean dosage per kg body weight (r = 0.4). Only in two cases did we notice therapeutic effects with serum levels below 200 mg/dl, corresponding to dosages of about 60-70 mg/kg, and frequently we saw toxic signs above serum levels of 300 mg/dl, corresponding to a dosage of more than 100 mg/kg. Therefore, the socalled therapeutic range can approximately be restricted to this area (60-100 mg/kg and 200-300 mg/dl BR in serum), yet wide interindividual differences have to be considered. Four of the five patients treated with BR monotherapy were responders. The rate of responders among the patients, who were on BR combined with other AEDs, was remarkably, though not statistically significantly, lower (n = 31 (56%)). When combined with BR, there was no AED with unequivocal advantages as to anti epileptic efficacy compared with other antiepileptics. Surprismgly, comparison of the efficacy of BR therapy
with the results of a therapy based on PHT and/or PB in the control group showed no important differences (Fig 1). The success rate of the BR therapy even exceeded the rate achieved in the control group by 10%, however, the difference was not significant (p < 0.05) (Table Sa). The rate of responders with a positive therapeutic effect for more than one year was also clearly, but not significantly(p = 0.05), higher in the BR group (Table 5b). Further, there was no significant difference with respect to the frequency of side effects or to the rate of drop-outs due to side effects, since in eleven cases in the BR group and ten in the control group the medication had to be discontinued due to adverse effects. Side effects of the BR therapy were noticed relatively frequently in about one half the cases (n = 32 (53%)). Mostly, however, they were tolerable or relieved on reduction of the dosage. Most predominantly, neurotoxic symptoms were combined with disturbances of the general condition, such as a loss of appetite and weight or, in single cases, nausea and vomiting or dryness of the mouth. Neurotoxic symptoms were psychic and motor slowness, apathy and drowsiness, muscular hypotonia and ataxia, tremor, indistinct speech and hypersalivation, along with delta wave slowing in the EEG in more severe cases. However, coma, dementia or delirious or psychotic syndromes were not observed. Dermatological side effects were of lesser importance: Acne was provoked three times and aggravated in two patients, and once a bromoderma occurred and necessitated discontinuation of the medication. Only in two cases were isolated psychic disturbances without contemporaneous neurotoxic symptoms were noticed_ In these patients, special neuropsychological investigations were not performed. Severe complications of the BR therapy did not occur. Emergency treatment with sodium-chloride infusions, diuresis or even hemodialysis was not necessary. With dosage reduction of discontinuation of the therapy in more severe cases, the side effects always disappeared rapidly and completely. The symptomatology of severe bromism, as described in the old literature [2], was not
Steinhoff and Kruse: Bromide therapy in intractable
ercs
147
observed. As mentioned above, BR had to be discontinued in eleven cases because of side effects: Once because of a bromoderma, twice because of worsening of the epilepsy in the case of minor seizures combined with GTCS, and eight times because of neuro-toxicity with a concurrent insufficient effect on GTCS. In four cases, worsening of the epilepsy occurred not due to a higher GTCS frequency, but to increased minor seizures: Once tonic, and three times myoclonic seizures were provoked, including a status myoclonicus in one patient, who had become free of GTCS. On the other hand, a positive BR effect on minor generalized and tonic seizures was also noticed in six and four patients, respectively, so that the increase in tonic and myoclonic seizures under BR therapy did not appear to be a specific complication among our patients. Neurotoxic symptoms occurred more frequently among non-responders than among responders. Cessation of seizures or essential improvement thus, as a rule, was not attained at the expense of serious adverse effects. There was no strict correlation of neurotoxic symptoms with either the dosage or SC of BR. A dosage of 50 mg/kg and SC of 200 mg/dl might come along with definite neurotoxic symptoms; dosages of more than 80 mg/kg and BR SCs of 250-300 mg/dl might be tolerated without side-effects. This indicates idividual variation in the tolerance of BR. In almost all patients with dosages of more than 100 mg/kg, corresponding with BR SC of more than 300 mg/dl, did neurotoxic symptoms occur, as mentioned previously. DISCUSSION Considering our results, there is an indication for the BR therapy, at least in cases of epilepsies with GTCS resistant to therapy with first-line AEDs, not only in children, but also in adolescents and young adults. Although our results were significantly better for patients younger than seven years, it is remarkable that ten of 39 patients older than seven years with drug-resistant epilepsies - sometimes for several years - became free of GTCS or responded very well. We noticed the lowest success rate, when there were minor generalized and/or localization-related seizures in addition to GTCS, and tonic seizures like in LennoxGastaut syndrome associated with GTCS. An increasing frequency of tonic seizures, as described in other reports [21,22,24], did not occur in our group except in one case. An increase in the frequency of myoclonic seizures, which was previously described as well [25], was observed in three cases. On the other hand, the frequencies of minor generalized and of tonic seizures were reduced in 10 (six and four, respectively) of 36 patients with either minor generalized or tonic seizures associated with GTCS,
148 Brain & Development, Vol 14, No 3, 1992
thus giving a success rate of more than 25%. The results of the present study suggest the conclusion that there is no contraindication of BR for definite seizure types so far they are pharmaco-resistant to conventional AEDs. There are great individual differences concerning both the antiepileptic efficacy and the frequency of side effects, depending on the BR dosage and the BR serum level. Considering these individual differences, we found that the therapeutic range was between about 200 and 300 mg/dl SC, which is about 1/3 higher than previously reported [20]. The reason for this might be that our determination method, that is, x·ray fluorescence analysis, results in higher values due to the improved accuracy, and due to the fact that organic bromine compounds are additionally detected and measured [2,28,29]. In fact, in a recently realized comparative study it could be demonstrated that the SCs of BR determined by a common colorimetric method are about 30% lower than determined by the x-ray fluorescence analysis [30]. We experienced five cases with BR monotherapy, four of which led to remarkable improvement, thus indicating that BR might be a successful AED not only as an addon-drug. We did not find a particularly good combination of BR with a definite AED like VPA, as previously reported [26] . This first clinical comparison of BR treatment with antiepileptic therapies based on first-line AEDs (PHT and/or PB) confirmed the high efficacy of BR as to GTCS, which was formerly assumed by several authors [14-18, 21-25]. In our group of patients suffering from severe epilepsies with very frequent GTCS, we found that both the therapeutic efficacy and the side effects of the BR therapy were comparable to those of PHT and/or PB. Conforming with numerous reports published in the last 130 years [2], the risk of irreversible defects induced by BR is extraordinarily low. In fact, we confirm again [31] that there is an indication for BR therapy in every case of pharmaco-resistant epilepsy with GTCS, presupposing conscientious monitoring consisting of regular neurological and psychiatriC investigations, and determinations of BRSC. REFERENCES
1. Locock C. Discussion of a paper by Sieveking, EH. Analysis of fifty-two cases of epilepsy observed by the author. Lancet 1857;1:528. 2. Steinhoff B. Zur antiepileptischen Therapie mit Bromiden: DarsteUung in der Literatur - Rontgenfluoreszenzanalytische Bestimmung von Bromiden in Korperfliissigkeiten .:. Klinische Ergebnisse der Behandlung von 60 Grand-mal-Epilepsien mit Bromiden (Medical thesis). Heidelberg: University of Heidelberg, 1989. 3. Ashmore Be. Comparative Luminal, Bromide, diet and eliminative treatment of epilepsy. Boston Med Surg J 1922; 137: 950-2.
4. Bennett AH. An inquiry into the effects of the prolonged administration of the bromides in epilepsy. Lancet 1884; 1:883. 5. Bondurant ED . Remarks upon the medicinal treatment of chronic epilepsy. Am J Insan 1894;51 :23-39. 6. Boshes B. Study of action of bromides in clinical and experimental epilepsy. J Nerv Ment Dis 1936;83: 390-404 . 7. Fere C. La bromuration ahautes doses dans l'epilepsie. Rev Med 1893;13:177-98. 8. McCallum AJ . The colony and bromide treatment of epilepsy. Br Med J 1903;1:616-8. 9. Pulliam JM. Epilepsy with special reference to treatment (discussion). JAMA 1914;67:252. 10. See J. Epilepsie et bromure; du bromure de potassium. Sem Med 1884;4:401-3 . 11. Wilks S. Bromide and iodide of potassium in epilepsy: cases and clinical remarks. Med Times Gaz 1861;2 :635-6. 12. Southerland RW. Relative effects of phenobarbital and sodium bromide as anticonvulsants in epileptic psychoses. Psychiatry Q 1940;14:382-7. 13. Berman W. The use of bromides for epilepsy. Am J Dis Child 1975;129:259-60. 14. Dreyer R. 1st Brom als Antiepileptikum noch zeitgemiiss? In: Deutsche Sektion der Internationalen Liga gegen Epilepsie, ed. Rundbrief 1971 ;Nr 41: 11. 15. Livingston S, Pearson SH. Bromides and the treatment of epilepsies in children. Am J Dis Child 1953;86:717-20. 16. Livingston S. Antiepilepticdrugs. AmJNurs 1963;63:103-7. 17 . Livingston S. Uber die Anwendung von Bromiden in der Behandlung der Epilepsien des Kindesalters. In: Deutsche Sektion der Internationalen Liga gegen Epilepsie, ed. Rundbrief 1971 ;Nr 41 :12-4. 18. Livingston S, Pauli LL. The use of bromides for epilepsy. Am J Dis Child 1975;29:259. 19. Matthes A. Epilepsien. 4th ed. Stuttgart · New York: Georg Thieme Verlag, 1984.
20 . Schmidt D. Behandlung der Epilepsien. 2nd ed. Stuttgart· New York : Georg Thieme Verlag, 1984. 21. Boenigk H-E, Lorenz J-H, Jiirgens U. Aktuelle Erfahrungen mit Bromiden zur Behandlung generalisierter Epilepsien. In: Kruse R, ed. Epilepsie 84. Reinbek: Einhorn Presse Verlag, 1984:316-25. 22 . Boenigk H-E, Lorenz J-H, Jiirgens U. Bromide - heute als antiepileptische Substanzen noch niitzlich? Nervenarzt 1985; 56:579-82. 23. Dreifuss FE, Bertram EH. Bromide therapy for intractable seizures (abstract). Epilepsia 1986;27:593. 24. Ernst J-P, Doose H, Baier WK. Bromides were effective in intractable epilepsy with generalized tonic-clonic seizures and onset in early childhood. Brain Dev (Tokyo) 1988; 10: 385-8. 25. Scheunemann W. Brom - eine therapeutische Chance bei friihkindlicher Grand-mal-Epilepsie. In: Hallen 0, MeyerWahl JG, Braun J, eds. Epilepsie 83. Reinbek: Einhorn Presse Verlag, 1983: 106-12. 26. Woody RC. Bromide therapy for pediatric seizure disorder intractable to other antiepileptic drugs. J Child Neurol 1990; 5:65-7. 27. Commission on classification and terminology of the International League against Epilepsy (lLAE). Proposal for classification of epilepsy and epileptic syndromes. Epilepsia 1985; 26:268-78. 28. Hahn-Weinheimer P, Hirner A, Weber-Diefenbach K. Grundlagen und praktische Anwendung der Rontgenfluoreszenzanalyse (RFA). Braunschweig·Wiesbaden: Friedrich Vieweg und Sohn,1984 . 29. Steinhoff B, Kruse R. Zur antiepileptischen Therapie mit Bromiden - eine klinische Studie. In: Wolf P, ed. Epilepsie 88. Reinbek: Einhorn Presse Verlag, 1988:374-8. 30. Kruse R. Personal communication, 1991. 31. Steinhoff B, Kruse R. Bromide therapy of tonic-clonic epilepsies: a clinical study (abstract). Epilepsia 1990 ;31: 225-6.
Steinhoff and Kruse : Bromide therapy in intractable CTCS
149
Bromide Treatment of Pharmaco-Resistant Epilepsies with Generalized Tonic-Clonic Seizures: A Clinical Study Bernhard J Steinhoff, MD and Rolf Kruse, MD
In a retrospective controlled clinical study we investigated the efficacy of bromides (BR) as to the frequency of generalized tonic-clonic seizures (GTCS) as a mono therapy (n = 5) or combined with other antiepileptic drugs (AEDs) (n = 55), and compared the results statistically with those achieved with a treatment based on phenobarbitone (PB) and/or phenytoin (PHT) in a very similar group of another 60 patients. All patients treated with BR had previously proved to be pharmaco-resistant. In most of the cases the epilepsies were based on early cerebral lesions, and were partly associated with severe mental and physical handicaps. Most of the patients were between three and 14 years old. The percentage of responders was 58%, which means a reduction in the GTCS-frequency of more than 50%. The positive result was observed for 28 months on the average. 27% of all patients became free ofGTCS for 18 months on the average (range, 44 days to 62 months), and 32% were improved (reduction in GTCS-frequency of more than 50% while not GTCS-free) for nearly three years on the average (range, 79 days to 110 months). Compared to PHT or PB, BR were slightly more efficient and similarly tolerable, but no statistically significant difference was seen (48% responders in the control group versus 58% responders in the BR-group). Thus, the statistical co"elation with competitive therapies confirmed the good efficacy of BR against GTCS, which was previously demonstrated by clinical tests without control studies. We recommended BR in the case of every pharmaco-resistant epilepsy with GTCS as a therapeutic alternative. Key words: Epilepsy, pharmaco-resistance, treatment, bromides, phenobarbitone, phenytoin, serum concentrations. Steinhoff BJ, Kruse R. Bromide treatment of pharmaco-resistant epilepsies with generalized tonic-clonic seizures: a clinical study. BrainDev 1992;14:144-9
The bromides (BR), introduced as antiepileptic drugs (AEDs) by Locock in 1857 [1], had lost their clinical importance much more since the appearance of phenobarbitone (PB) in 1912, phenytoin (PHT) in 1938 and then modern first-line AEDs [2]. However, their efficacy, especially in the case of generalized tonic-clonic seizures (GTCS), has been emphasized in several, though noncontrolled and statistically not satisfactory, investigations and reports [3-11]. Still in 1940, for instance, Southerland [12] reported that 85% of 40 patients treated with BR in addition to PB had less seizures than with PB alone,
From the Department of Children and Adolescents, Epilepsy Center Kork, Kehl-Kork. Received for publication: December 3, 1991. Accepted for publication: February 6,1992. Correspondence address: Dr. med. Bernhard J. Steinhoff, Neurologische Klinik, Ludwig-Maximi1ians-Universitiit, Marchioninistr. 15, D-8000 Miinchen 70, Germany.
and thus recommended BR as a very effective add-onAED. For 55 years BR had been the only effective AEDs. Therefore their antiepileptic indication was unlimited. Later, BR were recommended for the treatment of therapy-resistant epilepsies with GTCS in children and adolescents by only a few authors [13-18]. Livingston [17] and Berman [13] recommended them also for myoclonic seizures, and Matthes [19] for the treatment of partial seizures. Others [20] used BR only in cases of epilepsy with porphyria. Nevertheless, in some specialized epilepsy centers dealing with intractable epilepsies, BR were still used as effective add-on-drugs in selected cases, mainly with GTCS [2]. Recently, the clinical usefulness of BR was re-confirmed by several authors [21-26]. However, a statistical comparison of the clinical efficacy of BR with results achieved with established first-line AED has never been reported.
We report the results of a retrospective and controlled study concerning the efficacy of long-tenu BR therapy in 60 patients with pharmaco-resistant GTCS to conventional AED treatment.
Table 1 Clinical data
The efficacy of long-term BR therapy was investigated in 60 patients with pharmaco-resistant GTCS to conventional AEDs between 1972 and 1985. The control group also comprised 60 patients with GTCS treated with PB and/or PHT as a basic medication. For both the BR group and the controls the following selection criteria were used: 1. Epilepsies with either GTCS alone or combined with other seizure types. 2. Mean frequency of GTCS of two or more seizures per month for the two years before starting the treatment or from the very beginning of epileptic seizures in cases of a shorter duration than two years. 3. No or only slight changes in current antiepileptic co-medication. 4. No additionally introduced AEDs during the observation period. 5. Treatment only in Epilepsy Centre Kork. The definition of the patients with pharmaco-resistant GTCS to conventional AED treatment is those sufferred from severe epilepsies with GTCS which were insufficiently controlled by treatment with conventional AEDs like PB, primidone (PRM), PHT, carbamazepine (CBZ) or valproic acid (VPA) at the highest tolerable doses by Epilepsy Centre Kork. The clinical data are shown in Table 1. Both groups contained equal numbers of male and female patients, however, there were some, but not considerable, differences with respect to the mean ages of onset of epilepsy and of treatment, in addition to the mean duration of epilepsy, and to the relationship of pure and combined GTCS epilepsies, but not with respect to epileptic syndromes defined according to the proposal on the classification of epilepsies and epileptic syndromes of the International League against Epilepsy (ILAE) [27], and not with respect to etiological factors or to neurological and mental defects. In both groups the epilepsies were mostly severe residual syndromes based on early acquired cerebral lesions. In nearly all cases the epilepsy started in the first two years of life. Mental and neurological handicaps as well as pathological neuroradiological findings were commonly seen in these two groups. The group of generalized epilepsies consisted of symptomatic forms or both symptomatic and/or idiopathic fonus (groups 2.2 and 2.3 in the ILAE classification) including Lennox-Gastaut syndrome. Two patients in the control group and only one in the BR group had an
Control group
30/30 1.3 (0.3-11.3) (2.2) 10.0 (1.5-27.6) (5.6) 8.9 (0.6-25.9) (5.5)
30/30 2.3 (0.2-12.4) (3.0) 8.9 (0.5-18.1) (4.1) 6.3 (0.1-17.7) (4.2)
20 40 21
18 42 9
49
46 2 12 21 39
n = 60
Gender male/female Mean age at onset of epilepsy Range (years)
PATIENTS AND METHODS
Bromide group
(s)
Mean age at onset of treatment Range (years) (s)
Mean duration of epilepsy Range (years) (s)
Epilepsy syndrome Generalized Undetermined Epilepsy with GTCS alone Etiology Residual Progressive * Unknown Neurological deficits Mental defects Monotherapy
26 44 5
Combined therapy
55
11
n = 60
PB PHT PB PHT PB+PHT
4 5 18 17 16
* In both cases tuberous sclerosis, GTCS: generalized tonic-clonic seizures, PB: phenobarbitone, PHT: phenytoin. idiopathic generalized epilepsy (group 2.1. in the classification), and there was no case of typical localizationrelated symptomatic or idiopathic epilepsy (group 1 in the classification). The majority had to be classified as undetermined epilepsies (group 3 in the claSSification) with both generalized and focal signs, and seizures like generalized epilepsies beginning as idiopathic or symptomatic epilepsies with GTCS, but later in the course with partial seizures like complex-partial seizures additionally and with corresponding focal abnormalities in the EEG. There were no cases of severe myoclonic epilepsy in infancy, ESES syndrome or Landau-Kleffner syndrome in either group. If GTCS was not the only seizure type, there were additional tonic, minor generalized and localization-related seizures in different combinations and frequencies. Five patients were treated with BR monotherapy, while five patients were only on PHT and four on PB in the control group. BR were combined with a single AED (PB, PRM, PHT, CBZ, or VP A) in 20 cases and with two or more other AEDs in 35 cases, reflecting the severity of these epilepsies with GTCS of high frequency. 57 patients received KBr and the others CaBr2. Only BR serum concentration (SC) values determined by x-ray fluorescence analysis were examined. This
Steinhoff and Kruse: Bromide therapy in intractable
ercs
145
Table 2 Success rate of bromide treatment Effect Free of GTCS Improvement (n) >50% (n)
Duration of therapy effect > > > >
1.5 months 3 months 1 year 3 years
16 12 6 3
Responders (%)
Total no of patients
58% 55% 59% 67%
60 55 32 15
19 18
13 7
GTCS: generalized tonic-clonic seizures, improvement> 50%: reduction in GTCS-frequency > 50%, responders: free of GTCS/ improvement> 50%.
Table 3 Age at onset of bromide therpay and success rate (n = 60) Age (years)
n
FreeofGTCS (n)
Improvement >50% (n)
Responders
14
13
3
(%)
54%] 31%
46%*
* p < 0.05, GTCS: generalized tonic-clonic seizures, improvement> 50%: reduction in GTCS-frequency > 50%, responders: free of GTCS/improvement > 50%.
method, described elsewhere in detail [17, 34], was applied in 49 cases. The aim of the present study was to determine the therapeutical effects on the GTCS-frequency and the side effects of the BR-treatment compared with the control therapies. The effectiveness was classified as follows: 1. Free of GTCS for at least 1 month, i.e., the two fold time of being seizure-free before starting the treatment with BR or with PHT or PB in the control group. 2. Improvement, i.e., a reduction in GTCS-frequency by more than 50%, for at least the same time. These two subgroups are also summarized as "responders." 3. No effect. 4. Increase in GTCS-frequency, i.e., worsening. We compared the results achieved in both groups by means of the chi 2 -test, modified according to Yates (p = 0.05) [2]. RESULTS Table 2 shows that 16 patients treated with BR became free of GTCS and 19 patients improved remarkably. Therefore, 58% of the patients were responders for at least six weeks, when they had attained the BR steady state. This duration of at least six weeks is three fold
146 Brain & Development, Vol 14, No 3, 1992
longer than the response duration which had been attained during the two years before the BR treatment was started. Beside this initial positive effect the mean response duration was remarkably long: Patients, who had become GTCS-free, remained seizure-free for 18 months on average (44 days to 62 months). Improvement lasted even for 36 months on average (79 days to 110 months). In the follow-up period the success rate remained between 55% and 67% (Table 2). Thus, a loss of efficacy was not experienced. In one third (n = 12) of the 35 responders GTCS reappeared at the frequency seen in the pretreatment period. However, if the positive effect was maintained for longer than one year, we noticed a relapse of the pretreatment GTCS-frequency in only four of the remaining 19 responders. Many correlations were examined, i.e., between the success rate and different parameters of supposed prognostic importance, but only a few correlations were of statistical significance, for instance, the correlation of the success rate and age at the onset of BR therapy (Table 3). The treatment of patients younger than seven years, who showed more generalized epilepsies than average, proved significantly more often to be successful than the treatment of older patients. Although we did not find any statistically significant correlation between the duration of epilepsy and the success rate, it is remarkable that a positive therapeutical effect, sometimes for years, was obtained in half the 41 patients with a duration of epilepsy of more than five years (n = 22) and in 1/3 of the 25 cases with epilepsy for more than ten years (n = 9). With respect to the classification of the epileptic syndromes [27], a positive therapeutical effect could be obtained only in 55% of the cases of undetermined epilepsies (n = 22), compared to 65% of those of generalized epilepsies (n = 13). However, the rate of a positive long-term effect for more than one year was significantly higher in the cases of undetermined epilepsies (n = 16 (73%) versus n = 3 (23%), p 1.5 months
48% NS
Bromide group Control group
60 60
b. Success rate after> 1 year
66%
Bromide group Control group
Responders: free of generalized tonic-clonic seizures or improvement> 50%, NS: not significant (p < 0.05).
32 29
Responders: free of generalized tonic-clonic seizures or improvement> 50%, NS: not significant (p = 0.05).
30 Number of patients (n ~ , 201
2 ·· .. ·3·
free of GTCS -
improvement **
Bromide Therapy
(n = 60)
I!l!II
no effect
wo rsening
Control Therapy
(n = 60)
Fig 1 Efficacy as the CTCS* in bromide and control groups. *generalized tonic·c1onic seizures, **reduction in CTCSfrequency> 50%.
The mean SC determined in 49 patients was 253 mg/dl (156 mg/dl to 367 mg/dl). More than 70% of the measured mean SCs were between 200 and 300 mg/dl. The incidence of responders with mean SCs higher than 250 mg/dl was higher than with lower SCs, but again the difference was not significant (Table 4). We found a weak correlation between SC and the mean dosage per kg body weight (r = 0.4). Only in two cases did we notice therapeutic effects with serum levels below 200 mg/dl, corresponding to dosages of about 60-70 mg/kg, and frequently we saw toxic signs above serum levels of 300 mg/dl, corresponding to a dosage of more than 100 mg/kg. Therefore, the socalled therapeutic range can approximately be restricted to this area (60-100 mg/kg and 200-300 mg/dl BR in serum), yet wide interindividual differences have to be considered. Four of the five patients treated with BR monotherapy were responders. The rate of responders among the patients, who were on BR combined with other AEDs, was remarkably, though not statistically significantly, lower (n = 31 (56%)). When combined with BR, there was no AED with unequivocal advantages as to anti epileptic efficacy compared with other antiepileptics. Surprismgly, comparison of the efficacy of BR therapy
with the results of a therapy based on PHT and/or PB in the control group showed no important differences (Fig 1). The success rate of the BR therapy even exceeded the rate achieved in the control group by 10%, however, the difference was not significant (p < 0.05) (Table Sa). The rate of responders with a positive therapeutic effect for more than one year was also clearly, but not significantly(p = 0.05), higher in the BR group (Table 5b). Further, there was no significant difference with respect to the frequency of side effects or to the rate of drop-outs due to side effects, since in eleven cases in the BR group and ten in the control group the medication had to be discontinued due to adverse effects. Side effects of the BR therapy were noticed relatively frequently in about one half the cases (n = 32 (53%)). Mostly, however, they were tolerable or relieved on reduction of the dosage. Most predominantly, neurotoxic symptoms were combined with disturbances of the general condition, such as a loss of appetite and weight or, in single cases, nausea and vomiting or dryness of the mouth. Neurotoxic symptoms were psychic and motor slowness, apathy and drowsiness, muscular hypotonia and ataxia, tremor, indistinct speech and hypersalivation, along with delta wave slowing in the EEG in more severe cases. However, coma, dementia or delirious or psychotic syndromes were not observed. Dermatological side effects were of lesser importance: Acne was provoked three times and aggravated in two patients, and once a bromoderma occurred and necessitated discontinuation of the medication. Only in two cases were isolated psychic disturbances without contemporaneous neurotoxic symptoms were noticed_ In these patients, special neuropsychological investigations were not performed. Severe complications of the BR therapy did not occur. Emergency treatment with sodium-chloride infusions, diuresis or even hemodialysis was not necessary. With dosage reduction of discontinuation of the therapy in more severe cases, the side effects always disappeared rapidly and completely. The symptomatology of severe bromism, as described in the old literature [2], was not
Steinhoff and Kruse: Bromide therapy in intractable
ercs
147
observed. As mentioned above, BR had to be discontinued in eleven cases because of side effects: Once because of a bromoderma, twice because of worsening of the epilepsy in the case of minor seizures combined with GTCS, and eight times because of neuro-toxicity with a concurrent insufficient effect on GTCS. In four cases, worsening of the epilepsy occurred not due to a higher GTCS frequency, but to increased minor seizures: Once tonic, and three times myoclonic seizures were provoked, including a status myoclonicus in one patient, who had become free of GTCS. On the other hand, a positive BR effect on minor generalized and tonic seizures was also noticed in six and four patients, respectively, so that the increase in tonic and myoclonic seizures under BR therapy did not appear to be a specific complication among our patients. Neurotoxic symptoms occurred more frequently among non-responders than among responders. Cessation of seizures or essential improvement thus, as a rule, was not attained at the expense of serious adverse effects. There was no strict correlation of neurotoxic symptoms with either the dosage or SC of BR. A dosage of 50 mg/kg and SC of 200 mg/dl might come along with definite neurotoxic symptoms; dosages of more than 80 mg/kg and BR SCs of 250-300 mg/dl might be tolerated without side-effects. This indicates idividual variation in the tolerance of BR. In almost all patients with dosages of more than 100 mg/kg, corresponding with BR SC of more than 300 mg/dl, did neurotoxic symptoms occur, as mentioned previously. DISCUSSION Considering our results, there is an indication for the BR therapy, at least in cases of epilepsies with GTCS resistant to therapy with first-line AEDs, not only in children, but also in adolescents and young adults. Although our results were significantly better for patients younger than seven years, it is remarkable that ten of 39 patients older than seven years with drug-resistant epilepsies - sometimes for several years - became free of GTCS or responded very well. We noticed the lowest success rate, when there were minor generalized and/or localization-related seizures in addition to GTCS, and tonic seizures like in LennoxGastaut syndrome associated with GTCS. An increasing frequency of tonic seizures, as described in other reports [21,22,24], did not occur in our group except in one case. An increase in the frequency of myoclonic seizures, which was previously described as well [25], was observed in three cases. On the other hand, the frequencies of minor generalized and of tonic seizures were reduced in 10 (six and four, respectively) of 36 patients with either minor generalized or tonic seizures associated with GTCS,
148 Brain & Development, Vol 14, No 3, 1992
thus giving a success rate of more than 25%. The results of the present study suggest the conclusion that there is no contraindication of BR for definite seizure types so far they are pharmaco-resistant to conventional AEDs. There are great individual differences concerning both the antiepileptic efficacy and the frequency of side effects, depending on the BR dosage and the BR serum level. Considering these individual differences, we found that the therapeutic range was between about 200 and 300 mg/dl SC, which is about 1/3 higher than previously reported [20]. The reason for this might be that our determination method, that is, x·ray fluorescence analysis, results in higher values due to the improved accuracy, and due to the fact that organic bromine compounds are additionally detected and measured [2,28,29]. In fact, in a recently realized comparative study it could be demonstrated that the SCs of BR determined by a common colorimetric method are about 30% lower than determined by the x-ray fluorescence analysis [30]. We experienced five cases with BR monotherapy, four of which led to remarkable improvement, thus indicating that BR might be a successful AED not only as an addon-drug. We did not find a particularly good combination of BR with a definite AED like VPA, as previously reported [26] . This first clinical comparison of BR treatment with antiepileptic therapies based on first-line AEDs (PHT and/or PB) confirmed the high efficacy of BR as to GTCS, which was formerly assumed by several authors [14-18, 21-25]. In our group of patients suffering from severe epilepsies with very frequent GTCS, we found that both the therapeutic efficacy and the side effects of the BR therapy were comparable to those of PHT and/or PB. Conforming with numerous reports published in the last 130 years [2], the risk of irreversible defects induced by BR is extraordinarily low. In fact, we confirm again [31] that there is an indication for BR therapy in every case of pharmaco-resistant epilepsy with GTCS, presupposing conscientious monitoring consisting of regular neurological and psychiatriC investigations, and determinations of BRSC. REFERENCES
1. Locock C. Discussion of a paper by Sieveking, EH. Analysis of fifty-two cases of epilepsy observed by the author. Lancet 1857;1:528. 2. Steinhoff B. Zur antiepileptischen Therapie mit Bromiden: DarsteUung in der Literatur - Rontgenfluoreszenzanalytische Bestimmung von Bromiden in Korperfliissigkeiten .:. Klinische Ergebnisse der Behandlung von 60 Grand-mal-Epilepsien mit Bromiden (Medical thesis). Heidelberg: University of Heidelberg, 1989. 3. Ashmore Be. Comparative Luminal, Bromide, diet and eliminative treatment of epilepsy. Boston Med Surg J 1922; 137: 950-2.
4. Bennett AH. An inquiry into the effects of the prolonged administration of the bromides in epilepsy. Lancet 1884; 1:883. 5. Bondurant ED . Remarks upon the medicinal treatment of chronic epilepsy. Am J Insan 1894;51 :23-39. 6. Boshes B. Study of action of bromides in clinical and experimental epilepsy. J Nerv Ment Dis 1936;83: 390-404 . 7. Fere C. La bromuration ahautes doses dans l'epilepsie. Rev Med 1893;13:177-98. 8. McCallum AJ . The colony and bromide treatment of epilepsy. Br Med J 1903;1:616-8. 9. Pulliam JM. Epilepsy with special reference to treatment (discussion). JAMA 1914;67:252. 10. See J. Epilepsie et bromure; du bromure de potassium. Sem Med 1884;4:401-3 . 11. Wilks S. Bromide and iodide of potassium in epilepsy: cases and clinical remarks. Med Times Gaz 1861;2 :635-6. 12. Southerland RW. Relative effects of phenobarbital and sodium bromide as anticonvulsants in epileptic psychoses. Psychiatry Q 1940;14:382-7. 13. Berman W. The use of bromides for epilepsy. Am J Dis Child 1975;129:259-60. 14. Dreyer R. 1st Brom als Antiepileptikum noch zeitgemiiss? In: Deutsche Sektion der Internationalen Liga gegen Epilepsie, ed. Rundbrief 1971 ;Nr 41: 11. 15. Livingston S, Pearson SH. Bromides and the treatment of epilepsies in children. Am J Dis Child 1953;86:717-20. 16. Livingston S. Antiepilepticdrugs. AmJNurs 1963;63:103-7. 17 . Livingston S. Uber die Anwendung von Bromiden in der Behandlung der Epilepsien des Kindesalters. In: Deutsche Sektion der Internationalen Liga gegen Epilepsie, ed. Rundbrief 1971 ;Nr 41 :12-4. 18. Livingston S, Pauli LL. The use of bromides for epilepsy. Am J Dis Child 1975;29:259. 19. Matthes A. Epilepsien. 4th ed. Stuttgart · New York: Georg Thieme Verlag, 1984.
20 . Schmidt D. Behandlung der Epilepsien. 2nd ed. Stuttgart· New York : Georg Thieme Verlag, 1984. 21. Boenigk H-E, Lorenz J-H, Jiirgens U. Aktuelle Erfahrungen mit Bromiden zur Behandlung generalisierter Epilepsien. In: Kruse R, ed. Epilepsie 84. Reinbek: Einhorn Presse Verlag, 1984:316-25. 22 . Boenigk H-E, Lorenz J-H, Jiirgens U. Bromide - heute als antiepileptische Substanzen noch niitzlich? Nervenarzt 1985; 56:579-82. 23. Dreifuss FE, Bertram EH. Bromide therapy for intractable seizures (abstract). Epilepsia 1986;27:593. 24. Ernst J-P, Doose H, Baier WK. Bromides were effective in intractable epilepsy with generalized tonic-clonic seizures and onset in early childhood. Brain Dev (Tokyo) 1988; 10: 385-8. 25. Scheunemann W. Brom - eine therapeutische Chance bei friihkindlicher Grand-mal-Epilepsie. In: Hallen 0, MeyerWahl JG, Braun J, eds. Epilepsie 83. Reinbek: Einhorn Presse Verlag, 1983: 106-12. 26. Woody RC. Bromide therapy for pediatric seizure disorder intractable to other antiepileptic drugs. J Child Neurol 1990; 5:65-7. 27. Commission on classification and terminology of the International League against Epilepsy (lLAE). Proposal for classification of epilepsy and epileptic syndromes. Epilepsia 1985; 26:268-78. 28. Hahn-Weinheimer P, Hirner A, Weber-Diefenbach K. Grundlagen und praktische Anwendung der Rontgenfluoreszenzanalyse (RFA). Braunschweig·Wiesbaden: Friedrich Vieweg und Sohn,1984 . 29. Steinhoff B, Kruse R. Zur antiepileptischen Therapie mit Bromiden - eine klinische Studie. In: Wolf P, ed. Epilepsie 88. Reinbek: Einhorn Presse Verlag, 1988:374-8. 30. Kruse R. Personal communication, 1991. 31. Steinhoff B, Kruse R. Bromide therapy of tonic-clonic epilepsies: a clinical study (abstract). Epilepsia 1990 ;31: 225-6.
Steinhoff and Kruse : Bromide therapy in intractable CTCS
149
